Better to use a metal version for the front panel ESD plug. It ensures good contact with the metal plate, and "pigtail" connection can be then avoided.
Info: The chosen starpoint is qualified for 50Hz
Starpoint of earthing cables could be moved to the side wall of the chassis. Today it's on a back panel, which complicates removal of that panel (however, in practice this panel is not meant to be removed once the crate is assembled).
Chassis structure looks good in terms of shielding -> if properly done the chassis can take the full shielding function which results in lower criticality for cables inside
For the ribbon cables inside the crate, the closer they are to the metal plate the better
BNC/Coax looks like it is isolated to the chassis which may lead to coupling of external disturbances through relays to inside the chassis -> Better to connect outer BNC to chassis
Regarding the usage of two backplanes (main + 2x power) no particularly negative effect is seen
Fan tray connector: shielded cable is recommended which is to be connected with the front panel. The shorter this cable is, less noise it will pick up.
In addition it would be beneficial to ensure that the cables (especially ribbon cables) won't touch the relay
If the fan tray is not connected in operation there is the possibility of coupling to the backplane through the ribbon cable -> therefore and in general it is advised to guide the cables inside close to the chassis (e.g. close to the metal plate that divides the front-panel space from the PSU space).
For the 2 ethernet connectors it would be good to use a component with finger contact for a good connection to the front panel
Also a good connection of the front panel to a carrier front panel should be ensured to keep the path (connector->finger contacts->front panel->chassis) as short as possible - FMC EMC gaskets that are already used are good solution to that.
Good practice to verify the often fine gaps between the chassis polygon and other power/GND polygons. During tests often sparks are seen between one and another.
In general keep paths as short as possible and route currents away to the front panel as soon as possible
Electromagnetic gaskets to be used for the front panel connection
Regarding the front panel and the SUB-D connector the same applies as for the Profinet FMC - SUB-D connector should be well touching an FMC front panel
Fan tray PCB:
SUB-D connector to be connected to the chassis
Radtol SB: (schematics not thoroughly pre-studied by Daniel)
Nothing particular detected on the design
Nonrad SB: (schematics not thoroughly pre-studied by Daniel)
No problem seen for the SFP cage which has contact fingers and usually a good connection
It should be made use of the existing mechanical anchors of the FMC mounting for a good chassis connection
The lower point of the FMC screw connection (next to the "DI/OT" writing) can be improved for a better connection with the chassis polygon
Very good that existing SI simulations have been performed
No particular review yet performed, but it has been good to collaborate closely with SY-EPC
Already performed individual EMC tests can or should be repeated with the ongoing new design